Speech Recognition Pronunciation Adaptation via Expected Response Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech recognition systems face challenges in accurately adapting pronunciations for individual users, particularly when multiple pronunciations are acceptable, leading to recognition errors and inefficiencies, especially in hands-free communication scenarios like inventory management systems.
Innovation Solution
The system adapts speech recognition pronunciations by using expected responses to update phoneme sequences in a pronunciation dictionary, comparing recognition hypotheses with expected responses, and storing and updating phoneme sequences based on occurrence counts and thresholds to improve accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If speech recognition systems use a fixed pronunciation dictionary, then the system is simple to operate, but it cannot adapt to user-specific pronunciation patterns leading to recognition errors
Solution Approach 1:
The speech recognition system automatically adapts to user-specific pronunciation patterns by monitoring recognition results and updating the pronunciation dictionary without requiring manual configuration. The system serves itself by learning from user interactions and autonomously adjusting its pronunciation models.
Solution Approach 2:
The system dynamically changes the pronunciation dictionary parameters based on observed user behavior patterns. By updating the phoneme sequences and pronunciation probabilities in the dictionary based on actual recognition results, the system adapts to individual user preferences while maintaining the overall system structure.
2Measurement precision
If the system requires repeated utterances for accurate recognition, then recognition accuracy can be improved, but user productivity decreases
Solution Approach 1:
The system uses feedback from recognition results to continuously improve accuracy. By monitoring whether recognized words match expected responses and using this feedback to update the pronunciation dictionary, the system achieves higher accuracy without requiring repeated utterances, thereby maintaining user productivity.
3Adaptability or versatility
If the system updates pronunciation dictionaries frequently, then adaptability to user patterns improves, but network traffic increases
Solution Approach 1:
The system updates the pronunciation dictionary selectively based on observed user behavior patterns rather than continuously. By triggering updates only when significant changes in user pronunciation patterns are detected, the system achieves adaptability while minimizing unnecessary network traffic and energy consumption.
Data Source
AI summary
Embodiments of the disclosure provide for adapting word pronunciations in a speech recognition system to a user(s) based on expected responses. Some embodiments receive input speech and generate a recognition hypothesis for the input speech. The search algorithm may be informed by a pronunciation dictionary. The recognition hypothesis may comprise a sequence of one or more words. Some embodiments compare the recognition hypothesis with at least one expected response to determine if the recognition hypothesis matches the at least one expected response. Some embodiments generate a phoneme sequence for each word in the recognition hypothesis. Some embodiments after determining that the recognition hypothesis matches the at least one expected response, update a set of phoneme sequences in the pronunciation dictionary associated with at least one word of the recognition hypothesis.